Event choreography vs orchestration in Kafka - Practice Questions

Challenge - 5 Problems

❓ Predict Output

intermediate

Output of event choreography with Kafka consumers

Consider a Kafka setup where multiple microservices consume and produce events independently without a central coordinator. What will be the output sequence of events if Service A produces an event, and Services B and C react to it by producing their own events?

Kafka

Service A produces event 'start'.
Service B listens for 'start' and produces 'B_done'.
Service C listens for 'start' and produces 'C_done'.
No central orchestrator is involved.

A Events 'B_done' and 'C_done' are produced before 'start'.

B Events in order: 'start', 'B_done', 'C_done' or 'start', 'C_done', 'B_done' depending on timing.

C Only 'start' event is produced; 'B_done' and 'C_done' are never produced.

D Events in order: 'start', 'B_done', 'C_done' always in this fixed order.

💡 Hint

Think about how independent consumers react to events without a central controller.

In event choreography, services react independently to events. The order of 'B_done' and 'C_done' depends on which service processes the 'start' event first, so both orders are possible.

❓ Predict Output

intermediate

Output of event orchestration with Kafka

In an event orchestration model using Kafka, a central orchestrator service listens for events and commands other services. If the orchestrator receives 'start' and then commands Service B and Service C sequentially, what is the expected order of events?

Kafka

Orchestrator receives 'start'.
Orchestrator sends command to Service B.
Service B produces 'B_done'.
Orchestrator waits for 'B_done' before commanding Service C.
Service C produces 'C_done'.

A Events in order: 'start', 'B_done', 'C_done' always.

B Only 'start' event is produced; no other events follow.

C Events 'B_done' and 'C_done' produced simultaneously after 'start'.

D Events in order: 'start', 'C_done', 'B_done' always.

💡 Hint

Consider how the orchestrator controls the flow by waiting for events before proceeding.

The orchestrator commands Service B first and waits for 'B_done' before commanding Service C, so the events occur in a strict order.

🔧 Debug

advanced

Identify the error in Kafka event orchestration code snippet

Given the following Kafka consumer code snippet for an orchestrator, identify the error that would cause the orchestration to fail.

Kafka

def orchestrator_consumer():
    for message in kafka_consumer:
        if message.key == 'start':
            kafka_producer.send('service_b_topic', b'command')
        elif message.key == 'B_done':
            kafka_producer.send('service_c_topic', b'command')
        elif message.key == 'C_done':
            print('Process complete')
        else:
            print('Unknown event')

A RuntimeError because kafka_producer is not defined.

B Logical error: orchestrator sends commands in wrong order.

C No error; code runs correctly.

D SyntaxError due to missing colon after else statement.

💡 Hint

Check the syntax of conditional statements carefully.

The else statement is missing a colon at the end, causing a SyntaxError.

🧠 Conceptual

advanced

Difference in failure handling between choreography and orchestration

In Kafka-based microservices, how does failure handling typically differ between event choreography and event orchestration?

A Both choreography and orchestration use the same failure handling mechanisms.

B Choreography uses centralized error handling; orchestration relies on local retries only.

C Choreography relies on local retries and eventual consistency; orchestration can implement centralized error handling and compensation.

D Choreography does not handle failures; orchestration ignores failures.

💡 Hint

Think about who controls the flow and error management in each model.

In choreography, each service handles its own errors and retries, leading to eventual consistency. Orchestration allows a central controller to manage errors and compensations.

❓ Predict Output

expert

Number of events produced in mixed choreography and orchestration

In a Kafka system, an orchestrator sends a 'start' command to Service A. Service A produces an event 'A_done' which triggers Service B and Service C independently (choreography). Service B produces 'B_done' and Service C produces 'C_done'. How many total events are produced including the initial command and all reactions?

Kafka

Orchestrator -> 'start' command
Service A listens to 'start', produces 'A_done'
Services B and C listen to 'A_done' and produce 'B_done' and 'C_done' respectively

A 4 events: 'start', 'A_done', 'B_done', 'C_done'

B 3 events: 'start', 'B_done', 'C_done'

C 5 events: 'start', 'A_done', 'B_done', 'C_done', 'orchestrator_done'

D 2 events: 'start' and 'A_done' only

💡 Hint

Count all events produced by orchestrator and services.

The orchestrator produces 'start', Service A produces 'A_done', and Services B and C produce 'B_done' and 'C_done', totaling 4 events.